Meeting Notes 20090504 Face 2 face chat with Melissa Haendel on OBO Foundry practices

From Nemo

Jump to: navigation, search

Meeting notes with Melissa Haendel -- May 4, 2009

These are a few notes that I took away from my discussion with Melissa Haendel regarding her experiences working on ontology development with the OBO Foundry and ZFIN. She has a PhD in biology and specializes in anatomy aspects of zebrafish at ZFIN. She is the lead on the CARO and Zebrafish Anatomy ontologies. Here are the take-away points I got:

  • you MUST start using a version control repository especially when your group goes beyond 2 or 3, email won't cut it
  • avoid editing the text (or Word document) directly unless you're expert at it like Chris Mungall, because you will cause yourself unnecessary grief with syntax problems. Try not to use the Word document idea... that is too much work and unusable by the community.
  • choose what you like: OBO-Edit or Protege. It will help you decide a lot of things. Biologists prefer OBO because it "doesn't have a lot of things they don't need."
  • take a look at Tortoise SVN to make using CVS and SVN easy
  • follow the OBO Foundry best practices ("How to Join")
  • work closely with BIRN to determine how you will integrate with their ontologies (which are large and complex)
  • you can load multiple ontologies at once in OBO-Edit! the XREF tags are what "align" them. The way to do it is to setup a new load profile with the several urls and you can save the profile for later.
  • register an OBO prefix with Chris Mungall. unlike in OWL, namespace prefixes are GLOBALLY defined and registered for OBO.
  • check out OBO Foundry "inter-ontology" links wiki and "naming terms" (re: importing, cross-referencing terms, sub-classing remote terms)
  • use temporary placeholders for def and synonym citations (e.g. CARO:MAH, NEMO:GAF)
  • link permanent citations to actual online references (e.g., PUBMED IDs, or ZFIN:ZDB-PUB-10101-10292)
  • the OBO prefix refers to a URL, the ID refers to a page or resource at that URL, so the DBXREF tells you EXACTLY where a specific definition came from (i.e., what paper), the DBXREF should essentially like directly to it in this abbreviated syntax
  • ask yourself this question seriously: why do I need an ontology? (can you do without them? what makes reasoning necessary? even with some reasoning, can you do without an ontology?) idea: ontology development is a HUGE investment, without a lot of funding available to do it, so make sure they are worth it. KNOW: *what* reasoning you want, deal with *how* later.
  • have Gwen come talk to Melissa next time she's in town, to get a hands-on sense for the tools and processes for coordinating the ontology development with remote groups! She'd be happy to give a tour.

(Melissa is literally down the hall from Als' office at the NIC.)

/wiki/images/1/17/Fish1.png /wiki/images/e/ea/Fish2.png /wiki/images/f/fa/Fish3.png /wiki/images/f/ff/Fish4.png /wiki/images/4/40/Fish5.png /wiki/images/c/c5/Fish6.png
Personal tools